The mine migrant labour system: origins, recruitment, foreign labour and stabilisation¶
Was entrenched migrancy an economic necessity or a political construct, how did recruitment and monopsony work, and why did the mines shift away from foreign labour after 1974?
Works engaging the debate: 50
Positions¶
- Political bargain, not economic necessity (Jeeves and Crush)
- Low-grade ore cost constraint (Johnstone)
- Internalisation as deliberate policy (Clarke, ILO) vs response to supply shocks (Crush, Lucas)
- Effects on labour-sending economies (Lucas, Cobbe); monopsony vs elastic supply
Evidence base¶
The debate has been argued mainly from aggregate employer statistics. The core series is the Chamber of Mines, WNLA, NRC and TEBA count of black mine employees by territory of origin, usually annual or year-end. Several works rest on it. Yudelman and Jeeves (1986a) and Yudelman and Jeeves (1986b) append a 1920-83 table of source-area shares. Harington et al. (2004) traces some twenty source territories from 1896 to 1996 using the TEBA recruitment database and the Chamber's Statistical Tables. Crush (1986b) and Crush (1986c) use TEBA regional recruit counts for 1975-83 and the foreign/domestic composition series. Crush and James (1995a) and Crush and James (1995b) add mine-level and mining-house employment by source area for 1986-92. Wage evidence is almost always taken second-hand from Wilson's 1972 gold-mine earnings series, as in Harvey (1974), (Knight (1976a)), (Lemon (1984a)), (Lucas (1985)) and Yudelman and Jeeves (1986a). The historical and political arguments rest on archival correspondence. Jeeves (1975a), (Jeeves (1975b)), (Jeeves (1987)) and Jeeves and Crush (1992) draw on Native Affairs (NTS/SNA), GNLB and MNW files, PRO Colonial Office series (CO 879, CO 525), Zimbabwe CNC S1561 and Botswana Secretariat files, and Barlow Rand/CMIC and Rand Mines company letters. Taylor (1978), (Murray (1995)) and Pirie (1993) use similar material. These works also draw on commission evidence: the Industrial Commission of 1897, the Transvaal Labour Commission of 1903, the Low Grade Mines Commission of 1919-20, the Lansdown Commission of 1943, and the Tomlinson and Fagan evidence. Single-industry case studies rest on company archives, as with TGME in Bonner and Shapiro (1987), and on NTS enquiry files, as with the Natal token system in Edgecombe (1987). James (1987) reads Transvaal Chamber of Mines annual series for 1936-50 on labour supply, output, costs and rations.
A smaller quantitative strand uses different data. Lucas (1985) estimates sectoral labour demand and supply by three-stage least squares on unnamed annual series for 1946-80; the overall supply equation has only 26 observations. Lucas (1987a) and Lucas (1987b) estimate migration, crop, cattle and estate-labour equations on country time series for 1946-78 covering Botswana, Lesotho, Malawi, Mozambique and South Africa, with 13 to 31 observations per equation. Nattrass (1976b) applies indirect standard-population estimation to the 1936-70 censuses and obtains about 1.49 million male oscillating migrants in 1970. Knight (1976a) and Knight (1976b) reconcile the 1946-70 censuses, employer returns and the agricultural census under alternative undercount assumptions. Evidence from sending areas is more varied. Taylor (1990) uses TEBA Gaborone Schedule A forms for 1982-85, village-level recruitment for 1976 and 1985, and the 1981 census for 183 settlements. Chirwa (1996) uses TEBA Lilongwe negotiating records from 1980 and the Pryor-Chipeta estate employment series for 1964-84. Cobbe (1986) uses Central Bank of Lesotho macro data. Crush and James (1995a) reports OLS regressions on twelve annual observations of Lesotho national accounts for 1980-91. Dinkelman and Mariotti (2014) stands apart: it combines WNLA station locations from 1937 with the 1977 and 1998 Malawi censuses in a difference-in-differences design on 480 district-age-sex cells.
Contested evidence¶
On origins, the two sides read the same Chamber-generated commission evidence in opposite ways. Jeeves (1975a) and Jeeves (1975b) take Reyersbach's 1907 yield evidence, the 1897 Industrial Commission testimony and Frankel's return calculations as support for a low-grade ore cost constraint. Yudelman and Jeeves (1986a) argues that the Chamber's submissions to the 1919-20, 1922, 1930-32 and 1943 inquiries were self-serving and have misled historians. Jeeves and Crush (1992) and Crush and James (1995a) make the political-bargain case from counts of stabilised workers. These include the 1913 NTS married-quarters survey, fewer than 1,500 workers in married quarters by 1931, the state's 3 per cent cap on Free State stabilisation, and 0.7 per cent of workers in married quarters in a 1974 survey of forty mines. The Roth reviews of the 1991 book (Roth (1995a), (Roth (1995b))) cite the Chamber's Tomlinson costing as showing family housing was unviable. That costing is itself employer evidence, and the reviews never test it.
On the post-1974 shift, the foreign-share series is read three ways. Harington et al. (2004) treats it as evidence that the internalisation policy succeeded. Crush (1986b) and Crush (1986c) argue the 1973-77 fall from 79.6 to 49.3 per cent tracks the Malawian (1974) and Mozambican (1976) withdrawals in their timing, and that the share was flat at about 40 per cent after 1977. Lucas (1985) infers a deliberate post-1974 preference from a dummy-variable coefficient worth roughly 70,000 workers. Crush and James (1995a) then finds that retrenchments in 1986-92 fell disproportionately on South Africans. The figures themselves do not agree. Crush (1986b) gives 79.6 per cent foreign in 1973, and Yudelman and Jeeves (1986a) gives 22 per cent South African in 1973. Kanyenze (2004) gives 72 per cent foreign by 1970. Harington et al. (2004) gives only 37 per cent foreign in 1966. Hindson (1978), relaying Clarke, counts 443,000 foreign Africans in all South African employment in 1970. These differences point to different denominators: gold only, all Chamber mines, or the whole economy. The debate rarely reconciles them.
On monopsony, (Lucas (1985)) and Lucas (1987a) estimate an upward-sloping, inelastic supply curve and reject the labour-reserve model. Yudelman and Jeeves (1986a), citing Gemmill's 1970 internal briefing, has the Chamber itself believing that foreign supply was effectively infinitely elastic. Knight (1976a) attributes falling South African recruitment to monopsony and foreign sourcing, not exhaustion of the labour pool. For the earlier period, (Jeeves (1975a)) shows WNLA's 1906 monopsony securing only 2,337 of 22,601 northern Transvaal recruits. Turrell (1986) caps independent contractors at about 20 per cent of supply. Murray (1995) argues that illicit recruiting persisted after the 1920s. On effects in sending countries, the evidence diverges sharply. The Lesotho multiplier is below one in the twelve-observation regressions of Crush and James (1995a), but 1.25-1.85 in the unshown model of Cobbe (1986). Lucas (1987a) reads the Malawi ban as raising estate employment almost 80 per cent by 1978. Chirwa (1996) instead shows estate and mine employment rising together in 1968-74.
Underused and missing sources¶
The most obvious neglected source is the WNLA/TEBA collection described in TEBA Numerical Filing System WNLA... (2006). It holds routine labour returns and statistics (file 38/9), wastage, mortality and desertion statements (files 17 and 38/10-13), wage and deferred-pay files including gold payment to Mozambicans (93/5), Board of Management minutes, and a WNLA 'native census' of recruiting areas (135/16). None of the debate's quantitative works cites it item by item. Mine-by-mine personnel, wage and housing records and the TEBA Manpower Data Centre files are repeatedly named as the test for claims about managerial intent in Crush and James (1995a), (Crush (1986b)) and Yudelman and Jeeves (1986a). They would settle the internalisation-versus-supply-shock question, but they go unused. Company cost and grade data that could test the low-grade thesis directly are also missing from Jeeves (1975a). So are Chamber internal policy minutes, Portuguese archives on Mozambican supply (flagged by Harries (1986)), and household remittance surveys in Lesotho, Malawi and Mozambique. Lucas (1985) and Lucas (1987a) do not itemise their series, so their estimates cannot be replicated.
Strength of key claims¶
| Basis | Share of key claims |
|---|---|
| Systematic data | 11% |
| Partial data | 59% |
| Illustrative or anecdotal | 18% |
| Asserted | 13% |
Comparative figures for all debates are on the Literature overview.
Best-evidenced works¶
Dinkelman and Mariotti (2014), Lucas (1985), Lucas (1987a), Yudelman and Jeeves (1986a), Taylor (1990), Crush and James (1995a)
Most-used primary sources¶
Holdings are matched by machine against the archive's file manifest; a dash means no match was found, not that the project certainly lacks the source.
Works and their positions¶
| Work | Position |
|---|---|
| Bonner and Shapiro (1987) | Holds that labour stability at TGME rested on coercion (flogging, tax drives, chiefly authority) but also on contrived consent: freedom to desert, relaxed compounds, access to land, women and beer. |
| Bonner and Shapiro (1987) | Treats Pilgrim's Rest as only a partial company town (no scrip or company store) whose owner dominated local officials, and argues that central government left TGME largely alone. |
| Chirwa (1996) | Rejects the three speculative explanations (labour for estates, Frelimo and isolation, personal pride). Stresses the wider socio-economic context: estate demand, peasant distress and a possible regional-political calculation. Shows the estates could have obtained their labour without the ban. |
| Chirwa (1996) | Disputes the view that estate growth simply drew labour home. Out-migration and estate employment rose together, which points to peasant-sector distress as the push factor. |
| Chirwa (1996) | Weaker parties have limited power, but brute economic realities and the entrenched power of the mines prevail in the long run. Follows Crush. |
| Cobbe (1986) | Attributes the shift to concern about the reliability of foreign supplies (Mozambique, Malawi) together with domestic employment pressures. Higher wages attracted South African workers despite bantustan unemployment, and once the mines had a labour surplus they could limit the Lesotho contingent. |
| Cobbe (1986) | Treats Basotho migrants as a replaceable reserve whose threatened expulsion is Pretoria's most powerful leverage over Lesotho. He argues the rational course for Pretoria is to keep the threat credible but never carry it out. |
| Cobbe (1986) | Notes that the Chamber expected foreign workers to be less militant, but this appears incorrect for Lesotho: NUM and Lesotho unions are linked, and Basotho miners were sacked after the 1985 strikes. |
| Crush (1982a) | Stresses internal dynamics: chiefly appropriation of surplus labour in Lesotho, and the social division of labour generally. He calls explanations resting only on coercion or market demand 'fundamentally misleading'. |
| Crush (1986b) | Rejects it as oversimplified and not supported by the data. The 1974-77 shift was driven by Malawian and Mozambican withdrawals and then stabilized, with only gradual attrition in novice intake. |
| Crush (1986b) | Argues mining capital did not control its supply conduits in the mid-1970s, and that earlier scholarship attributed too much control and policy coherence to it. |
| Crush (1986b) | Credits the VRG system but says stabilization owes as much to the regional labour surplus and the lack of alternative employment as to industry policy. |
| Crush (1986c) | Rejects it as oversimplified and resting on a changing foreign:domestic ratio. Sees the 1970s shift as a supply-side shock and the 1980s as stability around 40% foreign, with only gradual attrition via novice allocation. |
| Crush (1986c) | Argues that internalization accounts, like some neo-Marxist accounts, overstate the control of mining capital. The industry in the mid-1970s scrambled for labour region-wide. |
| Crush (1986c) | Persistence reflects skills, state political leverage, divide-and-rule and 'playing the mix'. Stabilization succeeded largely because of favourable structural conditions (a labour surplus), not just policy. |
| Crush (1992) | Rejects a one-sided view. In South Africa's mines the technologies chiefly reinforce older coercion, though they are ambiguous because workers benefit from some services and so take part in their own monitoring. |
| Crush (1992) | Agrees with James, Harries and Moodie that total control was never realised. Argues electronic surveillance may bring management closer to it without replacing physical enclosure. |
| Crush (1992) | Sceptical. Surveillance may reduce pressure to provide alternatives to compounds, and stabilised workers are harder to monitor electronically. |
| Crush and James (1995a) | Jeeves and Crush hold that it was a political bargain of 1906-11 serving white supremacy, not a product of profitability concerns. |
| Crush and James (1995a) | Crush rejects this: the foreign share stabilised after 1980 and layoffs fell mainly on South Africans. |
| Crush and James (1995a) | Nattrass finds a fall in real revenue (gold price and ore grades) was the main culprit, not costs; Jourdan endorses this. |
| Crush and James (1995a) | Coplan and Thoahlane see a 'retrenchment industry' aimed at union-busting and cutting labour costs; Seidman and Nattrass stress gold price and productivity pressures. |
| Crush and James (1995a) | Cobbe finds weak evidence that migration raises domestic output; the GNP multiplier is under one, so a fall in migration need not reduce GDP. |
| Crush and James (1995a) | Crush and James argue for normalisation: treat miners as ordinary workers and ordinary immigrants. |
| Crush and James (1995b) | Jeeves and Crush hold that entrenchment 'should not be seen as the inevitable product of economic necessity'. They stress the 1906-11 political bargain and the segregationist settlement. |
| Crush and James (1995b) | Crush rejects the ILO's 'internalisation' thesis. Foreign share stabilised around 1980 and rose later, and foreign labour became discretionary rather than necessary. |
| Crush and James (1995b) | Nattrass attributes the crisis mainly to falling real revenue (price and grade), not rising costs. Coplan and Thoahlane argue retrenchment is partly a strategy to break the NUM and cut labour costs. |
| Crush and James (1995b) | Cobbe's naive regressions suggest a GNP multiplier below one, so migration lowers domestic GDP. This challenges the household-level view of Murray and Spiegel, though he calls his evidence suggestive only. |
| Crush and James (1995b) | The editors favour 'normalisation' of both South African and foreign miners over the status quo, expulsion, or South Africanisation. Several contributors warn that stabilisation harms foreign and rural-attached workers. |
| Crush and Namasasu (1985) | Rejects push-pull spatial explanations of migration in favour of analysis of the underlying structure of the regional labour system. |
| Crush and Namasasu (1985) | Argues that this view is inadequate and that development policy must be read through Lesotho's place in the regional migrant labour system. |
| Davies and Head (1995) | Argues that the decline is driven by industry restructuring and the breakdown of apartheid controls, independent of the new government's policy, rather than by populist policy. |
| Davies and Head (1995) | Argues against treating them separately; claims a multiplier link in which mine retrenchments fuel clandestine migration. |
| Davies and Head (1995) | Accepts their three-phase periodisation and extends it, identifying two phases since the mid-1970s: internalisation and stabilisation, then downsizing. |
| Davies and Head (1995) | Notes the literature's explanations (uncompetitive underground wages, militancy of urban workers, regional political leverage, diversification of supply) and treats these constraints as now weakened. |
| Dinkelman and Mariotti (2014) | Finds the income effect of migration and remittances dominates substitution, producing net positive schooling gains, but only where child labour is a poor substitute for adult male labour (non-estate districts). |
| Dinkelman and Mariotti (2014) | Argues the literature concentrates on South Africa and ignores the impacts of the mine labour system on northern neighbours, and claims to open this area; Lucas (1985, 1987) is noted as dealing only with national aggregates. |
| Dinkelman and Mariotti (2014) | Treats the WNLA monopsony, uniform wages and deferred pay as institutional facts, and notes the mines' later turn away from foreign supplies toward local labour. |
| Edgecombe (1987) | Shows that the Natal coal industry pursued stabilised, family-based labour in defiance of official migrant labour policy. Government tolerated this until the economics of the industry changed in the 1950s. |
| Edgecombe (1987) | Treats the token system as debt peonage, with a degree of compulsion and managerial collusion behind its formally voluntary repayments. |
| Harington et al. (2004) | Holds that it is entrenched, pervasive and permanent, not temporary, with serious social costs, and cites Wilson, Crush et al., James and Webster in support. |
| Harington et al. (2004) | Describes the Chamber's 1974 policy of replacing foreign labour with South African labour as largely achieved. It attributes part of the shift to decisions by Zambia and Malawi. |
| Harries (1986) | Jeeves (reported) stresses chiefs, merchants and politicians pushing workers onto the market. Harries says this ignores African workers' own role and that the Mozambican conquest explains the cheap labour supply. |
| Harvey (1974) | Follows Wilson that foreign recruitment let mine employers avoid raising wages and kept tighter control against unrest. He argues that industry has found ways to manage migrancy, which explains why business objects to it less than to colour-bar constraints. |
| Hindson (1978) | Dismisses the claim that higher income reduces African labour supply as a rationalisation for paying low wages. |
| Houghton and Dagut (1973) | Planning Council and Fagan Commission call migrancy wasteful but not immediately abolishable; extracts convey the argument for gradual stabilisation. |
| Houghton and Dagut (n.d.) | Presents the Chamber of Mines' belief that lower wages would lengthen stays, then the Mines Department and Labour Commission verdict that the cut was 'injudicious' and reduced supply. |
| Houghton and Dagut (n.d.) | Includes Milner's letters and Labour Commission findings that African supply was insufficient, with opposing views from labour and some mine leaders. |
| Indicus (1909) | Attributes the shortage mainly to low and reduced wages and to magnates' tactics, not to any lack of available labour; rejects the case for compulsion. |
| Indicus (1909) | Strongly opposed; regards indenture and compound schemes as 'slavery hidden under a thin veil of hypocrisy' and expects them to follow from the poor-grade ore economics. |
| James (1987) | Only partly. Stope-cleaning mechanization was limited and pre-dated the shortage; most mines preferred outward expansion (new fields) and did not mechanize, expecting labour supply to normalize. Managerial and supervisory intensification was the more common response. |
| James (1987) | The Chamber conducted a vigorous defence of cheap African labour in the 1940s, justified by a profit squeeze under a fixed gold price, and sought state subsidy for any concession. |
| Jeeves (1975a) | Partly dissents. The post-war wage and expansion decisions owed more to commitment to low-grade ore than to greed, though he admits the thirty-shilling wage was absurdly low. He also disputes Denoon's claim that there was no shortage by late 1905. |
| Jeeves (1975a) | Shows that industry opinion was divided on it in 1902 and that the proponents of higher wages won. |
| Jeeves (1975b) | Partly dissents: wage and expansion decisions owed less to simple greed and more to the commitment to low-grade ore, though the post-war 30s wage was admitted to be absurdly low. |
| Jeeves (1975b) | Notes that industry spokesmen publicly held to the target worker argument, but that agreement within the industry was far from complete, and the increase in wages won out. |
| Jeeves (1987) | Holds that Gemmill's tropical recruiting strategy and heavy investment in the migrant system made alternatives to migrancy and low wages harder for the Chamber to consider. |
| Jeeves (1987) | Mines took the pick of the fit labour while farms got the sick, weak and very young or old. WNLA's 1939 arrangements diverted clandestine migrants away from northern Transvaal and Zululand farmers. |
| Jeeves and Crush (1992) | Argues that its entrenchment was the result of a political bargain and a conjuncture of forces, not of economic necessity alone. |
| Jeeves and Crush (1992) | Takes a middle position. The system was an uneasy compromise in which workers' preferences mattered, but the authors stress its coercive character and the corporate and state choices that perpetuated it. |
| Jeeves and Crush (1992) | Treats the claim as self-serving. Workers were never consulted, and the industry used it to justify migrancy. |
| Kanyenze (2004) | Attributes it mainly to the political decisions of newly independent supplier states and to South African retrenchment after 1987. It also stresses the mines' own policy of stabilising the workforce with more South African workers. |
| Kanyenze (2004) | Cites ILO/SAMAT and Standing et al. that migration to the mines has 'stabilised' into career miners, with fewer TEBA recruiting stations and more contracting. It also stresses that supplier-state dependence on remittances persists. |
| Knight (1976a) | Attributes the rises mainly to concern about dependence on foreign labour (Malawi, Mozambique) and an attempt to recruit nationals. The gold price made them possible but was not their cause. Rejects the mining industry's view of a drying-up labour pool. |
| Knight (1976a) | Treats the backward-sloping supply curve as a widely held view and offers a Todaro-type probabilistic migration explanation of why Homeland underemployment can coexist with fewer South African miners. |
| Knight (1976b) | Says the rises were not initially a response to current shortage. The gold price made them possible but did not cause them. The main motive was reducing dependence on risky foreign labour by recruiting South African nationals. |
| Knight (1976b) | Notes the influential belief in backward-sloping supply curves. Prefers Todaro-type probabilistic migration reasoning to explain falling recruitment of nationals alongside rising Homeland underemployment. |
| Lemon (1984a) | Argues the companies' own recruitment policies constrain a free labour market, but that in the 1970s they shifted from foreign to domestic supply, higher wages and partial stabilisation. Anglo American pressed for policy change. |
| Lewis (1984a) | Wants more weight on relations between rural households and on unequal access to resources, moving away from explanations that stress only external forces. |
| Lucas (1985) | Rejects the notion of a vast labour reserve with a perfectly elastic supply. It finds supply upward-sloping and inelastic, and supports Lipton's view that the backward-bending "target worker" argument rationalised low wages. |
| Lucas (1985) | Argues that a deliberate preference against foreign dependence, prompted by political uncertainty (Malawi ban, FRELIMO, 1975 compound disturbances), explains the shift, not only supply-side events in sending countries. |
| Lucas (1987a) | Rejects the simple labour-surplus theory: wage employment reduces crop output in the short run. Adds a long-run capital-accumulation channel through migrants' remittances. |
| Lucas (1987a) | Finds short-run output losses but long-run productivity and cattle gains from remittances. Out-migration may even be a requirement of long-term development of traditional agriculture, except in Lesotho. |
| Lucas (1987a) | The supply of miners is not infinitely elastic and responds to relative wages. This justified centralized recruiting, which exploits monopsony power (his 1985a paper). |
| Lucas (1987a) | Forced registration raised mine migration before 1962. Its end reduced mine labour available to the Rand and raised plantation wages. |
| Lucas (1987b) | Rejects a simple labour-surplus view. Withdrawal for wage work reduces traditional crop output ceteris paribus, so labour has a positive shadow cost. |
| Lucas (1987b) | Argues they are not. Supply responds to wage differentials, and the mining houses used centralised recruiting to exploit monopsony power, as developed in Lucas (1985a). |
| Lucas (1987b) | Mixed. There are short-run output losses, but remittance-financed long-run gains in crop productivity and cattle, except in Lesotho, where erosion and neglect of crops are noted. Some out-migration may even be a requirement of long-term agricultural development. |
| Lucas (1987b) | Emigration raised estate and plantation labour costs. States responded with quotas (Nyasaland), the 1974 Malawian ban, forced labour (Mozambique) and the TEBA recruiting ban in White Farm Areas (South Africa). |
| Moodie (2005) | Argues the Chamber sacrificed productivity (piecework, incentives) to keep low-grade mines viable, against the wishes of many managers and even the NRC's Taberer. |
| Murray (1980) | Rejects the either/or framing; both dissolution of conjugal ties and persistence of agnatic structure are real, and both stem from the political economy of the labour reserve. |
| Murray (1995) | Endorses Jeeves' challenge to the view that the monopsony grew naturally from capital concentration. Stresses mine owners' dependence on touts and freelancers rather than their prescient class power. |
| Murray (1995) | Disagrees with Jeeves: it persisted into the 1930s-40s and shifted to other employers, with mining agents also implicated. |
| Murray (1995) | Rejects a rigid dichotomy. Coercion and voluntary market participation were intertwined, and illicit recruiting facilitated proletarianisation. |
| Nattrass (1976b) | Realist reformist: the system cannot be abolished at once and should be phased out over decades. She favours family housing, decentralization, education subsidies and homeland-controlled labour supply. |
| Olivier (1929) | Rejects the idleness charge, but accepts that Africans with access to land supply wage labour intermittently and respond weakly to higher wages. He explains this by access to land and limited wants. |
| Olivier (1929) | Argues that compulsion fails and that improving African agriculture itself produces a willing labour supply. |
| Pirie (1993) | Argues railways were important but constrained, and that their key role was to perpetuate migrancy by making it practical and affordable. This delayed a stable, permanently urbanised mine workforce. Treats the question of whether railways were essential as open. |
| Roth (1995a) | The book, via the Tomlinson Commission report, holds that a family-housing system would be far more expensive, so replacing migrancy would not have been economically viable. |
| Roth (1995b) | As reported, the book stresses migrancy's role in sustaining rural households through remittances and deferred pay, and the superior conditions and wages the Chamber provided. Roth notes the resulting pro-Chamber impression. |
| Roth (1995b) | As reported, the Chamber's Tomlinson Commission report showed that family housing would cost far more than compound housing, so replacing the system seemed not economically viable. |
| Sanderson (1961) | Argues that hut and labour taxes, together with administrative practice such as 'finding work' for defaulters, were a strong stimulus to wage migration. Migration had begun before taxation or pacification, however, so taxes were an additional rather than the sole cause. |
| Sanderson (1961) | Judges the disadvantages (social disruption, retarded African agriculture, low wages) to outweigh the advantages (a softer transition, a mobility that acted as a reserve price on labour). |
| Sanderson (1961) | Holds that Nyasaland lacked the staff, sanctions and inter-governmental co-operation needed to regulate migration, so passes and recruitment bans were largely ineffective. |
| Spandau (1980a) | Argues that wage-bill substitution alone does not explain mechanization; output per unit of operating time, maintenance costs, labour availability and labour-control problems matter, and isoquants are often inelastic to wage changes. |
| Spandau (1980a) | Rejects wage equal to marginal product; fixed recruitment and training costs mean firms pay to cut turnover and stabilize a core workforce, following Oi, Becker and Doeringer and Piore. |
| Spandau (1980a) | Holds that mining houses have at times preserved labour-intensive methods, partly for social reasons, but that the policy is now being abandoned; the government is criticised for not removing factors that favour capital intensity. |
| Spandau (1980b) | Rejects the neo-classical wage equals marginal product account; argues fixed employment costs and internal labour markets explain high wages, stabilization and training. |
| Spandau (1980b) | Treats mechanization as the product of technical, economic and social factors. It stresses output per unit of operating time and labour control, with the 1973 Witbank unrest named as a trigger, rather than wage costs alone. |
| Taylor (1978) | Accepts drought, pestilence, taxation and market depression as contributing factors but gives greater weight to administrative policy: lack of opposition to recruiting and failure to stimulate the domestic economy. |
| Taylor (1978) | The Protectorate government never actively encouraged migration as British Bechuanaland's officials did, but regulated recruiters while treating them as inevitable and desirable, so it facilitated recruitment by omission and by design (hut tax, compulsion, support for the Chamber of Mines). |
| Taylor (1990) | Takes the structural side: Botswana evidence shows a permanent, self-sustaining shift towards stabilized, internalized labour, while conceding that sceptics (Martin, Yudelman and Jeeves) point to forecasting errors and the political leverage of diversified supply. |
| Taylor (1990) | Argues the decline is primarily demand-side (mine policy), not due to rising domestic employment in Botswana. |
| Taylor (1990) | Argues that the 'capricious target worker' has given way to a highly regulated career worker, with mine managements controlling reengagement. |
| Turrell (1986) | Turrell partly disagrees with Jeeves. He accepts the detail of competitive recruiting but holds that competition does not show capital's weakness. In his view monopoly capital means competition at a different level, not its abolition. |
| Turrell (1986) | Turrell finds Jeeves's stress on recruitment policies such as cattle advances a poor balance against Bundy-style proletarianization. He argues that different forms of African incorporation into wage labour need explaining. |
| Yudelman and Jeeves (1986a) | Rejects the view that it was a state-imposed 'internalisation' policy (Clarke, ILO writers). It was a shift to nearer, less vulnerable sources after foreign frontiers closed, with a high gold price making higher wages affordable. |
| Yudelman and Jeeves (1986a) | Argues that Johnstone and others accepted industry-crafted data. Cost constraints matter, but the persistence of migrancy also owes much to colour-bar rigidities and to the expansionist possibilities at the edges of the labour frontiers. |
| Yudelman and Jeeves (1986a) | Sceptical. Unemployment is structural, but the diversification of sources remains valuable to state and capital, and the future is open. Foreign migrants will probably continue to be recruited. |
| Yudelman and Jeeves (1986b) | Partly accepts the low-grade ore argument (Johnstone) but says historians have uncritically accepted industry-crafted data and Chamber propaganda. Migrancy also owes much to colour-bar rigidities and to the expansionist possibilities of the labour frontiers. |
| Yudelman and Jeeves (1986b) | Rejects the state-imposed view (Clarke) as having 'very little truth'. It stresses the closing of frontiers and the vulnerability of foreign supply, with the state and industry mostly aligned. |
| Yudelman and Jeeves (1986b) | Sceptical of the ILO-type claim that internalisation is permanent. Expects continued two-tier migrancy and sourcing diversification, though stabilisation is rising and the NUM will shape foreign migrants' fate. |